532 SECRETORY AND EXCRETORY SYSTEMS 



sphaerraphides occur mixed with solitary crystals in the secondary 

 phloem of Quercus peduneulata, Celtis australis, Morus alba, Fagus 

 si/lratica, etc. The same remark applies to Betula verrucosa and 

 Alnus glutinosa : but in these crystal dust is found in addition. J. 

 Moller concludes, as the result of repeated examination of various 

 kinds of bark, that crystal dust, raphides and sphaerraphides are 

 invariably contained in thin-walled cells, whereas well-developed 

 solitary crystals occur principally in or near sclerotic cells. Moller 

 seeks to correlate this distribution with a difference in the velocity 

 of osmosis in the various types of crystal-cell. It is obvious, however, 

 that the relative thickness of the wall is only one among a number of 

 factors which influence the rapidity of osmosis, and consequently of 

 crystallisation, in any particular case. A factor of much more funda- 

 mental importance is the varying rate of general metabolic activity ; as 

 a matter of fact, Moller himself points out that sphaerraphides prepon- 

 derate in the primary extra-cambial tissues of young, actively-growing 

 internodes, whereas solitary crystals are more generally found at a later 

 stage, after the formation of periderm and other secondary tissues. 

 From the same point of view, it is quite comprehensible why sphaer- 

 raphides always appear in connection with the autumnal depletion 

 of deciduous leaves, a process which is certainly accompanied by very 

 active metabolic changes. Quite apart from such nutritive factors, the 

 specific constitution of the protoplasts of the crystal-sacs undoubtedly 

 often determines the mode of crystallisation ; the nature of this proto- 

 plasmic control is, however, at present entirely unknown. 



Crystal- sacs like the crystals themselves vary greatly in form. 

 Not infrequently their shape stands in some relation to that of the 

 crystals which they contain, though as a rule the connection is quite 

 an indirect one. Thus, sphaerraphides generally occur in isodiametric 

 cells, while raphide- bundles are contained in tubular sacs. The cor- 

 relation is most marked in the case of the elongated crystal-sacs of 

 many Monocotyledons (e.g. Pontederiaceae), where the w T all often fits 

 closely around the dart- or needle-like crystalline enclosure. According 

 to Eothert, mechanical distension of the wall by the growing crystal is 

 inconceivable. Any correlation in respect of shape or mode of growth 

 between cell-membrane and crystal must therefore be entirely due 

 to the control exercised over both by the living protoplast of the 

 crystal- sac. 



All crystal-sacs, of course, contain living protoplasm to begin with. 

 According to Johow and Fuchs, the raphide-sacs of many Monocotyledons 

 and Dicotyledons permanently retain their living contents, including 

 the nucleus. More often, however, fully-developed crystal-cells are 

 dead structures containing very little beside the actual crystals. Both 



